Now showing 1 - 10 of 22
  • Publication
    Creating and Characterising Electricity Load Profiles of Residential Buildings
    Intelligent planning, control and forecasting of electricity usage has become a vitally important element of the modern conception of the energy grid. Electricity smart-meters permit the sequential measurement of electricity usage at an aggregate level within a dwelling at regular time intervals. Electricity distributors or suppliers are interested in making general decisions that apply to large groups of customers, making it necessary to determine an appropriate electricity usage behaviour-based clustering of these data to determine appropriate aggregate load profiles. We perform a clustering of time series data associated with 3670 residential smart meters from an Irish customer behaviour trial and attempt to establish the relationship between the characteristics of each cluster based upon responses provided in an accompanying survey. Our analysis provides interesting insights into general electricity usage behaviours of residential consumers and the salient characteristics that affect those behaviours. Our characterisation of the usage profiles at a fine-granularity level and the resultant insights have the potential to improve the decisions made by distribution and supply companies, policy makers and other stakeholders, allowing them, for example, to optimise pricing, electricity usage, network investment strategies and to plan policies to best affect social behavior.
      462Scopus© Citations 5
  • Publication
    Program Optimisation with Dependency Injection
    (Springer, 2013-04) ;
    For many real-world problems, there exist non-deterministic heuristics which generate valid but possibly sub-optimal solutions. The program optimisation with dependency injection method, introduced here, allows such a heuristic to be placed under evolutionary control, allowing search for the optimum. Essentially, the heuristic is “fooled” into using a genome, supplied by a genetic algorithm, in place of the output of its random number generator. The method is demonstrated with generative heuristics in the domains of 3D design and communications network design. It is also used in novel approaches to genetic programming.
      496Scopus© Citations 3
  • Publication
    Learning to Sparsify Travelling Salesman Problem Instances
    In order to deal with the high development time of exact and approximation algorithms for NP-hard combinatorial optimisation problems and the high running time of exact solvers, deep learning techniques have been used in recent years as an end-to-end approach to find solutions. However, there are issues of representation, generalisation, complex architectures, interpretability of models for mathematical analysis etc. using deep learning techniques. As a compromise, machine learning can be used to improve the run time performance of exact algorithms in a matheuristics framework. In this paper, we use a pruning heuristic leveraging machine learning as a pre-processing step followed by an exact Integer Programming approach. We apply this approach to sparsify instances of the classical travelling salesman problem. Our approach learns which edges in the underlying graph are unlikely to belong to an optimal solution and removes them, thus sparsifying the graph and significantly reducing the number of decision variables. We use carefully selected features derived from linear programming relaxation, cutting planes exploration, minimum-weight spanning tree heuristics and various other local and statistical analysis of the graph. Our learning approach requires very little training data and is amenable to mathematical analysis. We demonstrate that our approach can reliably prune a large fraction of the variables in TSP instances from TSPLIB/MATILDA (>85%) while preserving most of the optimal tour edges. Our approach can successfully prune problem instances even if they lie outside the training distribution, resulting in small optimality gaps between the pruned and original problems in most cases. Using our learning technique, we discover novel heuristics for sparsifying TSP instances, that may be of independent interest for variants of the vehicle routing problem.
      25
  • Publication
    Engaging Business Students in Quantitative Skills Development
    (Australian Business Education Research Association, 2015-06) ;
    In this paper the complex problems of developing quantitative and analytical skills in undergraduate first year, first semester business students are addressed. An action research project detailing how first year business students perceive the relevance of data analysis and inferential statistics in light of the economic downturn and the challenges society faces is discussed. Students¿ attitudes were evaluated via an online survey consisting of both quantitative and qualitative responses. While two thirds of respondents do acknowledge the relevance of such a course for future business roles, it is shown that more work must be done to distinguish between why data analysis is relevant and how data analysis is performed. Also discussed are findings related to student learning, their intellectual development, and their motivation and expectations upon enrolling on the Data Analysis for Decision Makers (DADM) module. The challenges in teaching such a mandatory module to Business students are discussed and a pedagogical framework for promoting deeper student engagement through active learning, regular continuous assessment and technology are also examined.
      152
  • Publication
    A Decomposition Algorithm for the Ring Spur Assignment Problem
    This paper describes the ring spur assignment problem (RSAP), a new problem arising in the design of next generation networks. The RSAP complements the sonet ring assignment problem (SRAP). We describe the RSAP, positioning it in relation to problems previously addressed in the literature. We decompose the problem into two IP problems and describe a branch-and-cut decomposition heuristic algorithm suitable for solving problem instances in a reasonable time. We present promising computational results.
    Scopus© Citations 5  480
  • Publication
    Prediction of Forestry Planned End Products Using Dirichlet Regression and Neural Networks
    (Society of American Foresters, 2015-04-12) ; ; ;
    We describe a set of nonparametric and machine learning models to forecast the proportion of planned end products (PEP) that can be extracted from a forest compartment. We determine which forest crop attributes are significant in predicting the product proportions (of sawlog, pallet, stake, and pulp) based on an Irish data set supplied by Coillte, the Irish state forestry company. Dirichlet regression and neural networks are applied to predict the product proportions and evaluated against a multivariate multiple regression benchmark model. Based on predictive performance, the neural network performs slightly better in comparison to Dirichlet regression. However, assessing the model logic and taking account of user interpretation, the Dirichlet regression outperforms the neural network. Both models are also compared to an existing rule-based model used by Coillte. The nonparametric and machine learning techniques provided consistent reliable models to accurately predict the PEP proportions. The two proposed models extend the versatility of nonparametric and machine learning techniques to areas such as forestry.
    Scopus© Citations 7  408
  • Publication
    Probability density distributions for household air source heat pump electricity demand
    The Irish government is implementing policies to transition Ireland to a low carbon and environmentally sustainable economy by 2050. Ireland has sectoral targets of 600,000 installed heat pumps by 2030, currently roughly 28,000 are installed. Such a high target of heat pumps will not only have a significant effect on electricity demand but also on the management and operation of the grid. In this paper we explore the demand from homes heated by air source heat pumps using an innovative dataset from a field trial in Ireland. To assess the impact of large-scale adoption of heat pumps, this paper estimates the after diversity maximum demand per heat pump heated home. In particular we explore statistical distributions to best model coincident demand, and estimate after diversity maximum demand per home. We use the software package RStudio to model several different distributions. Based on goodness-of-fit statistics and criteria, a Gamma distribution is the best fit. We apply our methodology to data from a similar heat pump trial in the UK to complement our results.
    Scopus© Citations 5  186
  • Publication
    Sub-hour Unit Commitment MILP Model with Benchmark Problem Instances
    Power systems are operated to deliver electricity at minimum cost while adhering to operational and technical constraints. The introduction of smart grid technologies and renewable energy sources offers new challenges and opportunities for the efficient and reliable management of the grid. In this paper we focus on a Mixed Integer Programming sub-hour Unit Commitment model. We present analysis of computational results from a large set of problem instances based on the Irish system and show that problem instances with higher variability in net demand (after the integration of renewables) are more challenging to solve.
      367Scopus© Citations 1
  • Publication
    A Branch and Cut Algorithm for the Ring Spur Assignment Problem
    The Ring Spur Assignment Problem (RSAP) arises in the design of Next Generation Telecommunications Networks (NGNs) and has applications in location-allocation problems. The aim is to identify a minimum cost set of interconnected ring spurs. We seek to connect all nodes of the network either on a set of bounded disjoint local rings or by a single spur edge connected to a node on a local ring. Local rings are interconnected by a special ring called the tertiary ring. We show that the problem is NP-Hard and present an Integer Programming formulation with additional valid inequalities. We implement a branch-and-cut algorithm and present our conclusions with computational results.
      493Scopus© Citations 9
  • Publication
    Smart Meter Tariff Design to Minimise Wholesale Risk
    (Elsevier, 2016-06) ;
    Smart metering in electricity markets offers an opportunity to explore more diversetariff structures. In this article a Genetic Algorithm (GA) is used to design Time ofUse tariffs that minimise the wholesale risk to the supplier in residential markets.Residential demand and the System Marginal Price of Ireland's Single ElectricityMarket are simulated to estimate the wholesale risk associated with each tariff.
      374Scopus© Citations 1